I GC Evidence Paper Firms

EVIDENCE PAPER
FIRM CAPABILITIES AND

ECONOMIC GROWTH
Nicholas Bloom (Stanford and IGC), Gregory Fischer (LSE and
IGC), Imran Rasul (UCL and IGC), Andres Rodriguez-Clare (UC
Berkeley and IGC), Tavneet Suri (MIT and IGC), Christopher Udry
(Yale and IGC), Eric Verhoogen (Columbia and IGC), Christopher
Woodruff (Warwick and IGC) and Giulia Zane (LSE)
11 April, 2014
The International Growth Centre (IGC) aims to promote sustainable

growth in developing countries by providing demand-led policy
advice based on frontier research.
Evidence Paper: Firm Capabilities and Economic Growth
Contents
Executive Summary ................................................................................................................. 3
1.
Introduction ...................................................................................................................... 8
2.
Data.................................................................................................................................. 9
3.
Stylised Facts ................................................................................................................. 12
4.
Determinants of Low Labour Productivity ..................................................................... 25

4.1
Credit Market Frictions ........................................................................................... 27
4.2
Institutional Constraints .......................................................................................... 29
4.3
Infrastructure ........................................................................................................... 37
4.4
Skills and Training ................................................................................................... 38
4.5
Firm Clusters ........................................................................................................... 40
4.6
Markets for Inputs and Outputs ............................................................................. 40
5.
Resource Allocation ....................................................................................................... 42
6.
Conclusion ..................................................................................................................... 46
References ............................................................................................................................ 49
Page 2 of 54
Firm Capabilities and Economic Growth

Nicholas Bloom (Stanford and IGC), Gregory Fischer (LSE and IGC), Imran
Rasul (UCL and IGC), Andres Rodriguez-Clare (UC Berkeley and IGC),
Tavneet Suri (MIT and IGC), Christopher Udry (Yale and IGC), Eric Verhoogen
(Columbia and IGC), Christopher Woodruff (Warwick and IGC) and Giulia
Zane (LSE)1
Executive Summary
The IGC Firm Capabilities Research Programme pulls economists with a common
interest in firm capabilities together to focus on three core questions: (i) what are
the key proximate determinants of firm productivity? (ii) Where does the productive
capacity of firms originate? (iii) What are the barriers that prevent resources from
moving from unproductive firms and sectors to areas of higher productivity? In
addition to the focus on larger firms within the manufacturing and service sector,
small firms (and farms) are also examined to reflect the fact that the majority of
citizens in developing countries are not employed by large firms.
This evidence paper on firm capabilities is structured as follows. In Section 2 we
discuss available data sets on firms in developing countries. This serves to
emphasize how little we know about this sector and in this section we present some
ideas on how to expand and improve our evidence base in this area. In Section 3 we
present stylised facts about firms in developing countries. Section 4 therefore
examines the determinants of low labour productivity. Here we not only look at what
factors are associated with low labour productivity but also ask where differences in
productive capacity across firms come from. This, in turn, greatly constrains our
ability to design effective policies to encourage industrial development. In Section 5
we take on the broader question of resource allocation and examine what prevents
resources to move from unproductive sectors and firms to more productive areas.
Currently most of what we know comes from the few countries that have made their
industrial survey and censuses available to researchers. Firm data collected by third
1
Nicholas Bloom, Gregory Fischer, Imran Rasul, Andres Rodriguez-Clare, Tavneet Suri, Christopher
Udry, Eric Verhoogen and Christopher Woodruff are Directors of IGC Firm Capabilities Research
Programme; and Giulia Zane is PhD student at LSE. Imran and Gregory are grateful to Vittorio Bassi
(IFS) for excellent research assistance.
Page 3 of 54
parties such as the World Bank Enterprise Surveys (WBES) are also useful. There are
however important limitations of these different datasets. First, with the exception
of data from Mexico and India, most of the existing surveys only survey large
industrial firms. Second, many of the forces that we believe are important to a firms
capabilities are rarely measured in existing datasets. Third, we believe it is important
to measure entrepreneurship. Fourth, much of what we know about farms and small
informal firms comes from household surveys that do not collect much of the
information needed to look at barriers to productivity increase.
The IGC will support research that develops new methodologies to construct
appropriate measures of the drivers of firm productivity from existing data. For example,
the enterprise maps compiled by John Sutton and co-authors in Ethiopia, Ghana,
Tanzania, and Zambia are based on detailed original surveys of the fifty leading firms in
each of these countries. Specifically, Suttons research finds that few of the leading
firms in these countries originated as small firms; rather they are more likely to have
been started by entrepreneurs engaged as brokers in foreign trade. We want to
know whether this finding holds more systematically in other countries
The IGC Research Programme aims to find the explanations to some stylised facts
associated with developing countries. It has been widely observed that labour
productivity is several times greater in developed countries than in developing
countries. Explaining why this is the case, by delving into firm level data and by
collecting more firm level data to identify the determinants of low firm productivity,
represents a major objective of the Research Programme. It is also observed that
the size distribution of firms in most developing countries is heavily concentrated
around firms with relatively low levels of employment, even when one excludes sole
proprietors. Two immediate questions are raised by the presence of a much thicker
left-hand tail of the firm size distribution. First, why do low productivity firms not
improve their productivity? Second, why do markets not force the less productive
firms out of business?
We address the question of low productivity by examining information on the
constraints face by the firms in the WBES in this survey firms are asked to choose the
largest obstacle they believe they face from a list. Difficulty in accessing credit features
prominently in all IGC countries. Another obvious candidate is constraints related to
institutions. Indeed, in eight IGC countries, more entrepreneurs report institutional
constraints as being more important than those related to finance. Infrastructure is
seen as a major impediment to expansion and growth in most developing countries
Page 4 of 54
and this is true for both small and large firms. Another important constraint facing
firms in low income countries is the absence of adequately skilled workers and
managers. The structure of the market for inputs and outputs may also play an
important role in addition to its incentive effects on productivity. In the IGCs
previous work in Sierra Leone, Rwanda, and Pakistan, our research has shown that
the market structure of the agricultural supply chain has important effects on prices
and thus farmer welfare.
Moving beyond the material captured in available firm level datasets, we have
abundant evidence that the clustering of firms, particularly in industrial sectors,
produces positive externalities that are important for capturing local spillovers and
fostering industrial development. One of the important spillover effects of this is in
knowledge: firms may learn from their neighbours, either directly or by hiring
workers from nearby firms. Much of the evidence we have on where and why firm
clusters form in particular parts of developing countries is descriptive and
anecdotal. There is much more research that could be done in this important area.
The last pillar of the research on firm capabilities encompasses understanding the
importance of and barriers to effective resource allocation. There exists suggestive
evidence that growth is driven by the entry and emergence of new firms and sectors
as well as the reallocation of resources from less productive to more productive
firms. If resources do not flow into more productive new firms and sectors and if
there are impediments to resources being reallocated from unproductive firms and
sectors to more productive areas the economic growth is likely to be low for a
sustained period of time. The study of firm capabilities therefore has to encompass
the study of firm dynamics both in terms of entry and growth of new firms but also
in terms of whether more productive firms are growing and less productive ones
dying out.
We would like to see much more of type of work where the focus is on identifying
constraints on resource reallocation and trying to identify how these constraints can
be overcome. More broadly the IGC Firm Capabilities Programme would like to
understand the importance of the reallocation of resources from informal to formal
firms in manufacturing, from non-exporters to exporters, from small family owned
farms to large corporate farms, and from agriculture to industry. The reality in many
countries is that this necessary resource reallocation does not take place, resulting
in industrial sectors being dominated by a large number of small, unproductive,
informal firms. In the agricultural sector, the vast majority of land and people are
Page 5 of 54
employed in small subsistence farms, despite the presence of potentially more

productive farms producing cash crops with modern technologies. Furthermore,
workers and capital remain locked in less productive agricultural sectors, while it
may be the case that these resources could be more profitably employed in other
sectors. What are the barriers that prevent the efficient reallocation of resources?
What prevents subsistence farmers from moving into cash crops or into urban
industrial sectors?
Due to the research that the IGC funded earlier, we know that the magnitude of
internal trade barriers does have a potential effect on the growth of firms. Atkin and
Donaldson (2012) show that internal trade barriers may be very large in poor
countries. In the agricultural context, IGC-supported research suggests that
lowering transportation costs in Sierra Leone led to large gains among farmers and
also improved competition amongst traders by reducing search costs. In the
industrial sector, there is currently little evidence on whether internal trade costs
might have similar effects on firm productivity.
One key finding from this evidence paper is that what we know about the
production sector in developing countries is very incomplete. Part of the challenge
of the IGC Firm Capabilities Research Programme will therefore be to gather better
data on firm productivity across the full firm size distribution in developing countries.
By gathering this information over time and in a larger number of countries we will
be better able to study both the drivers of productivity as well as the resource
allocation and structural change processes that underpin economic growth. Our
review of the available data on firms in developing countries throws up three major
puzzles. The first is why the size of the industrial sector remains so small within the
overall economy? The second puzzle concerns why production is concentrated
within firms in the left hand tail of the productivity distribution? The third puzzle is
why more resources not reallocated from unproductive firms and sectors to
productive firms and sectors in developing economies?
Production in developing countries is concentrated in firms that are both small and
unproductive. This is true whether we look at agriculture, manufacturing or services.
The small size of the industrial sector, and the focus of production on unproductive
firms, helps to explain why aggregate productivity is so low in developing countries.
And low productivity ultimately explains why living standards are so much lower in
these countries. Understanding why firm capabilities are so low in developing and
uncovering ways both to increase labour productivity and to move resources from
Page 6 of 54
unproductive firms and sector to productive ones are therefore of fundamental

importance not just for understanding what drives economic growth in developing
countries but also for finding the means to close the gap in living standards
between developing and developed nations.
Page 7 of 54
1. Introduction
A well-established macro-economic literature2 argues that growth and employment
generation can be sustained only through increases in productivity at the aggregate
level. In turn, productivity at the level of the aggregate economy depends on the
underlying productivity of all firms in the economy. This is the case whether we are
looking at large formal firms, small informal firms, large commercial farms, or smallscale family farms. This raises three important questions:
1. What are the key proximate determinants of firm productivity?
2. Where does the productive capacity of firms come from?
3. What are the barriers that prevent resources from moving from unproductive
firms and sectors to areas of higher productivity?
Work under the IGC Firm Capabilities Research Programme will be focused on
these questions. In addition to our focus on larger firms within the manufacturing
and service sectors, these research questions reflect the fact that the majority of
citizens in developing countries are employed in unproductive small firms (including
farms), resulting in an urgent need to think through how to make these firms more
productive and also how to encourage the transition of workers into more
productive sectors.
The wide distribution of firms in developing countries implies that working on firm
capabilities in developing countries requires inputs from economists interested in
industrial development, trade, agricultural development and entrepreneurship. The
IGC Firm Capabilities Research Programme collects together the top economic
researchers in the world with these interests to work on the core questions
identified above. These researchers have a common interest in finding means of
increasing the productivity of firms at all levels in developing countries with a view
to increasing aggregate productivity in these countries.
This evidence paper on firm capabilities is structured as follows. In Section 2 we
discuss available data sets on firms in developing countries. This serves to
emphasize how little we know about this sector and in this section we present some
ideas on how expand and improve our evidence base in this area. In Section 3 we
present stylised facts about firms in developing countries. The key stylised fact that
emerges is that firms in are much more unproductive that their counterparts in the
2
Caselli (2004)
Page 8 of 54
developed world. Not only is the industrial sector in developing countries relatively
small but most production also tends to be carried out in firms that are both
unproductive and small. Informal firms in the manufacturing and service sectors and
subsistence farms are where the majority of the labour force is employed.
Low labour productivity not only constrains economic growth but also the
standards of living that an economy can support. Section 4 therefore examines the
determinants of low labour productivity. Here we not only look at what factors are
associated with low labour productivity but also ask where differences in productive
capacity across firms come from.
Our examination of the evidence here reveals our lack of knowledge of the factors
which underpin low productivity. This, in turn, greatly constrains our ability to design
effective policies to encourage industrial development. In Section 5 we take on the
broader question of resource allocation and examine what prevents resources to
move from unproductive sectors and firms to more productive areas. We see that
competition and trade are particularly important in determining whether resource
reallocations take place. What is clear is that entry of more productive firm and
sectors and the shift of resources from unproductive firms and sectors to
productive ones are both likely to be important. However, we are still at the
beginning of understanding what drives these two processes.
2. Data
We begin by discussing available data sources and our plans to develop new data
sources. Currently most of what we know comes from the few countries that have
made their industrial survey and censuses available to researchers. Firm data
collected by third parties such as the World Bank Enterprise Surveys (WBES) are
also useful. The countries that have made their data available include Chile,
Colombia, Mexico, Ethiopia, India, and China.
Although these surveys are large and have provided valuable information, they are
generally difficult to access. In contrast, the micro-data from the WBES is easily
accessible. The main limitation of the WBES data is that it only surveys registered
private firms with more than five employees. There are also questions about the
underlying sampling frame used in the WBES data. Nonetheless, the WBES data is
the only data readily available for most of the countries in the world. For example,
Page 9 of 54
they are available for 12 out of 14 IGC countries3 between 2006 and 2011. We will
therefore use this data later to present stylised facts on the private sector in the
IGC countries.
There are however important limitations of these different datasets. First, with the
exception of data from Mexico and India, most of the existing surveys only survey
large industrial firms. For example, the widely used Chilean industrial data only
covers firms with more than 10 employees. The WBES only covers registered firms
with more than five employees. This is problematic in many countries where
employment is in non-registered family firms. We are therefore missing a great deal
of information because our data does not survey firms in the part of the economy
that is central in understanding the constraints to growth in many countries. We will
therefore support efforts to measure informal firms, either through expanding
existing datasets or by making datasets with information on informal firms, such as
the data available from the Indian National Sample Survey and from the Mexican
Economic Census, more widely available. Understanding the workings of the
informal sector whether this be in small manufacturing or service sector firm or in
family run farms represents a major challenge going forward.
Second, many of the forces that we believe are important to a firms capabilities are
rarely measured in existing datasets. For example, although elements such as the
use of specific technologies and access to inputs are sometimes measured in
agricultural household surveys, such information is rarely collected in surveys of
industrial firms. Specifically, many of the most important determinants of a firms
capabilities such as knowledge, access to global production chains and input
markets, use of specific technologies, quality of products, contracting relationships
between different firms are rarely observed in datasets of industrial firms large
enough to permit statistical testing.
The IGC has supported original data gathering to address this limitation. For
example, the enterprise maps compiled by John Sutton and co-authors in
Ethiopia, Ghana, Tanzania, and Zambia are based on detailed original surveys of the
fifty leading firms in each of these countries. These maps provide useful thick
descriptions of the key sources of the productivity advantages of leading firms in
the IGC countries. Another example is IGC funded work by Bloom and co-authors
(2013) looking at the process of productivity improvements among textile firms in
India, and on-going work by Bandiera, Guiso, Prat and Sadun (2011) analysing the
3
All but Myanmar and South Sudan
Page 10 of 54
use of time by CEOs in India and other countries are examples of this type of work.
Such gathering of original data not available in existing datasets will continue to be
an important focus of our work in the future.
Third, we believe it is important to measure entrepreneurship. Specifically we want
to measure the extent to which entrepreneurs might have created firms in a different
institutional environment but did not do so because of the barriers they face in low
income countries. Obviously potential entrepreneurs are not measured in any
dataset because it is difficult to measure something that did not happen. The Global
Entrepreneurship Monitor Adult Population Survey (GEM) is a dataset that might be
useful in the future. The GEM is an individual-level survey designed to collect
information about entrepreneurial activity and aspirations. It covers five IGC
countries (Ghana, India, Pakistan, Uganda, and Zambia) and all surveys are
conducted between 2008 and 2010. As yet we have not utilised this survey because
the raw data are currently available only for India and Uganda.
Fourth, much of what we know about farms and small informal firms comes from
household surveys that do not collect much of the information needed to look at
barriers to productivity increase. There is a need to collect more information on
household firms and also to link household level data to firm level data where
household members work outside the household. This is true both because the
much production is household based but also because many individuals divide their
time between household based production and outside employment. Obtaining this
information over time on the same population of households and on the firms where
they are employed would be valuable as it would provide insights into the process
of structural change within an economy and what drive this process.
Fifth, we also need to develop new methodological approaches to measure
productivity at the firm level. Although the measurement of productivity at this level
is the subject of a large academic literature, the standard methods for estimating
productivity often assume away exactly the sort of heterogeneity that we believe
might be important in the countries the IGC operates in. The IGC will thus support
research that develops new methodologies to construct appropriate measures of
the drivers of firm productivity from existing data.
Page 11 of 54
3. Stylised Facts
As a motivation for the research programme, this section presents key stylised facts
about firms in poor countries. We begin with the key stylised fact we want to
explain, namely low labour productivity. Figure 14 displays labour productivity per
worker using data on firms with 10 to 49 workers from the World Bank Enterprise
Surveys. The median firm with 10 to 49 workers in El Salvador, Burundi, Nepal,
Afghanistan, Uganda and Mongolia generates less than US$4000 per worker in
annual sales. In comparison, in Mexico the median firm of this size generates
$17,332 per worker in sales, while firms in Croatia, Turkey and Brazil are generating
more than $50,000 per worker in sales.
Bloom et al. (2010) using accounting data likewise show that the average firm labour
productivity as measured by sales per employee is dramatically lower in developing
countries. This is shown in Table 15. Figure 1 and Table 1 display a great deal of
heterogeneity in labour productivity amongst developing countries but what is
perhaps is even more striking in Table 1 is the large gap in labour productivity
between developed and developing countries. It is this gap in productivity that
underpins differences in living standards between developed and developing
countries. Irrespective of the measure we use labour productivity is several times
greater in developed countries than in developing countries. Explaining why this is
the case, by delving into firm level data and by collecting more firm level data,
represents a major objective of the IGC Firm Capabilities Research Programme.
Source: World Bank Enterprise Surveys, data from Harrison et al. (forthcoming).
GDP per capita from the IMF 2005 in $PPP. Sales/Employee in current $, across all firms in the
ORBIS. Source: Bloom, Mahajan, McKenzie and Roberts (2013).
5
Page 12 of 54
Page 13 of 54
We only know a little about firm characteristics associated with low productivity. For
example, Bloom, Sadun and Van Reenen (2012) show that bad management
practices are associated with low firm productivity. A specific management practice
is the extent of decentralisation of decision-making. Bloom, Sadun and Van Reenen
(2012) use the international management practices data to examine patterns of
decentralisation across environments. Across countries and regions, they show that
decentralisation of decision-making is associated with greater trust and the rule of
law being better enforced. This correlation makes sense if we believe that wellfunctioning legal systems help to deal with contractual problems within firms. They
also find that decentralisation and the use of information technologies are
complementary with one another. This may be because information technology
helps firm owners to monitor firm managers. But it should be obvious that this
research only scratches the surface of the determinants of low firm productivity.
Yet, we note that in at least some low income countries, delegation does appear to
be possible. There is a large number of very large business groups involved in the
manufacture of garments sector in Bangladesh and in the manufacture of textiles in
India, for example. However, such organisations are the exception rather than the
norm in most low income countries. One way in which this fact may show up is in
the skewed distribution of firm productivity. Specifically, the size distribution of firms
in most developing countries is heavily concentrated around firms with relatively low
levels of employment, even when one excludes sole proprietors (see Hsieh and
Klenow 2009). For example, in Mauritius, out of the 91,980 manufacturing units with
nine or fewer workers, 70.6% have only one or two workers, and only 6.2% have
five to nine workers, while there are only 841 manufacturing firms in the whole
country with 10 or more workers.
Similar skewed distributions are found in other in a range of other developing
countries. In Tanzania there are only 80 manufacturing firms with 100 or more
workers, compared to 695 with 10 to 99 workers, and 24,204 with fewer than 10
workers6. In Sri Lanka, 92.8 percent of manufacturing establishments have fewer
than 10 workers. In Mexico and Bolivia, 91 percent of manufacturing establishments
employ fewer than ten workers, while less than 2 percent have 100 or more workers.
The percentage of microenterprises is even higher outside of manufacturing, with 97
McKenzie, 2011
Page 14 of 54
percent of retail firms and 94 percent of services firms in Mexico having fewer than
ten workers7.
When we consider firms with 10 or more workers, Table 2 shows that the majority of
these firms employs only 10 to 49 workers, while in most countries only 15 to 20
percent of firms with 10 or more workers have at least 100 workers. We also see
similar evidence from other studies. Bloom and Van Reenen (2012) show there is
substantially more dispersion in management practices in India and China than in
the United States and northern Europe. Hsieh and Klenow (2009) present evidence
showing more dispersion in total factor productivity in India and China, compared
with the United States. By either measure, we see a much thicker left-hand tail of
low-productivity firms surviving in China and India. What is also interesting is that,
relative to firms in developed countries, firms do not grow much with age in
developing countries (Hsieh and Klenow 2012). This suggests that firms, even if they
survive, are hindered from growing to scale in low income countries.
Pags, 2011
Page 15 of 54
Figure 2A presents similar data of the skewed size distribution for the IGC countries
from the WBES.
Figure 2A: Firm Size Distribution (Number of Employees), by IGC Country
As can be seen, the firm size distributions are quite similar across IGC countries.
The modal group consists of firms of size 0-9 full-time employees in nearly all IGC
countries, except Bangladesh and Ethiopia. If informal firms were included, we
would expect this pattern to hold throughout. It is also interesting to note that the
proportion of firms that are very large (with 100+ employees) varies across IGC
countries. For example, 28.1% of registered firms in Bangladesh are in this part of
the firm size distribution. In contrast, Liberia has almost no such firms. Indeed in
Liberia less than 1% of firms have more than 70 employees.
Figure 2B shows the distribution of firm size in the manufacturing vs. the services
sector. Given the skewed distribuiton of firm size shown in Figure 2A, this
Page 16 of 54
breakdown in sectoral composition focuses on firms with 0-49 full-time employees,

with the rightmost bar includes all firms with at least 50 full time employees.
Figure 2B: Sectoral Composition of Enterprises, by IGC Country
There are two points we take away from this figure. First, there appears to be
important differences in the share of the manufacturing sector versus the share of
the services sector between countries. In countries such as Bangladesh, Ghana,
Mozambique, Pakistan and Zambia, the formal sector appears to be dominated by
manufacturing firms (in all size categories). In countries such as Tanzania, Ethiopia,
Uganda, and Rwanda, formal firms appear to be largely in the service sector. So to
the extent that formal firms are important, it appears that such firms are missing in
the service sector in some countries and in the manufacturing sector in other
countries.
This pattern also shows up in other data sources. For example, The World Banks
Ethiopia at a Glance report indeed does seem to suggest that the manufacturing
Page 17 of 54
sector is indeed very small, with less that 6% of GDP arising from the
manusfacturing sectors in all reported years. In contrast, the service sector
accounts for around 40% of GDP in years close to the relevent WBES survey year in
Ethiopia.
Table 3: Structure of Ethiopian Economy, by Sector, 1991-2011
For Bangladesh, the 2001 Economic Census suggests that around 88% of
established firms are of size 0-4 and 62% are not registered. The manufacturing
sector is found to be the most important sector for establishments of size 10+, very
much in line with the WBES data.
Figure 2C examines the sectoral composition of larger firms, those with at least 100
full-time employees.
Page 18 of 54
Figure 2C: Sectoral Composition for Firms of Size 100+, by IGC Country
We see there is enormous variation in what larger firms are doing across IGC
countries. In some cases there is a contrast with the sectoral focus of smaller firms.
For example, in Tanzania the majority of large firms are in manufacturing, while the
majority of small firms are in the service sector.
The WBES only captures formally registered firms. However, this data still allow us
to explore some features of the informal sector. Specifically the WBES asks firms
whether they registered when they started business or after some time and whether
they have informal competitors. As Figures 3A and 3B show, this data suggests that
the majority of formally registered firms started out as formal registered firms or
registered soon after operations commenced. This evidence suggests little
movement from informality to formality.
Page 19 of 54
Figure 3A: Firms Formally Registered When Started
Page 20 of 54
Figure 3B: Number of Years Firms Operated Without Registration
Only a handful of countries make a concerted effort to identify and survey informal
firms.
The three countries we know of that survey informal firms are Mexicos Economic
Census (MEC), the survey of informal firms in the Indian National Sample Survey,
and the Indonesian Economic Census. Comparing the data from the Mexican
Economic Census with the distribution of firms in WBES sample for Mexico provides
some indication of the gaps in the WBES data. As shown in Figure 3C, small firms,
those with ten or fewer employees, are substantially more prevalent in the Mexican
Economic Census, representing over 90% of firms in manufacturing, retail and
services in contrast to roughly 40% in the weighted WBES data for the same
sectors. This is not a surprise. The WBES is expressly targeting registered firms with
five or more employees.
Page 21 of 54
Figure 3C: Sectoral Composition in Mexico, by Establishment Size
Yet as shown in Figure 3B these differences persist, albeit in smaller magnitude,

when restricting our attention to just firms with five or more employees in both
samples. The magnitude of informality, even in a middle income country such as
Mexico, means that the WBES data will yield a misleading picture of the aggregate
economy. Figure 3D shows that these differences extend to the measured sectoral
distribution, with the Mexican Economic Census over-representing both small retail
enterprises and, somewhat surprisingly, large manufacturers relative to the WBES.
The variation between these two survey methods underscores the potential
challenges of extrapolating from the experiences of formal firms to understand the
constraints facing the informal sector.8
Even the Mexican Economic Census data has its limitations. Data from the National Survey of
Employment and Occupations (ENOE) suggests that the census captures less than half of total
employment and still only 60% of private urban employment. Much of the activity not captured by
the census is informal (Busso, Fazio and Levy 2012).
Page 22 of 54
Figure 3D: Firm Size Distribution by Sector in Mexico, Excluding Establishments of

Size 0-5
We end this section by noting that we have little evidence on productivity dynamics
in developing countries. In high-income countries, data allow a detailed examination
of productivity dynamics. Using U.S. data, for example, Foster, Haltiwanger and
Krizan (2006) attribute substantially all of the impressive productivity gains in the
retail sector during the 1990s to entry of more productive firms and exit of less
productive firms. Using data from six narrowly defined manufacturing sectors,
Foster, Haltiwanger and Syverson (2008) find that the direct effect of entry explains
only around one-quarter of productivity growth. Most productivity growth comes
from improvements in incumbent firms.
There are no comparably comprehensive studies of productivity dynamics in lowincome countries, and this is something we want to correct. We note, however, that
the enterprise maps by John Sutton suggest that the origins of successful firms in
developing countries may be quite different than in more advanced countries.
Specifically, Suttons research finds that few of the leading firms in these countries
originated as small firms, rather they are more likely to have been started by
Page 23 of 54
entrepreneurs engaged as brokers in foreign trade. We want to know whether this

finding holds more systematically in other countries as if it is indeed the case, it
suggests that something is failing in the entrepreneurship process. This finding
would also be consistent with the evidence in Hsieh and Klenow (2012) which
shows that unlike firms in developed economies surviving firms in developing
economies do not show much of growing with age.
Specifically, this failure lies in that very few entrepreneurs manage to grow, almost
as if the equivalent of Google in Sub-Saharan Africa never grows beyond the
proverbial garage (or living room). We do not know what forces are behind this
pattern (assuming it holds up), but we suspect that access to output markets and
access to critical inputs such as capital, skills, imported intermediate goods, seeds,
and fertilizers (for farmers) may be critical. Understanding what the key forces are is
obviously critical, but at this time there is a lack of evidence on the importance of
most of these main suspects.
Page 24 of 54
4. Determinants of Low Labour Productivity

Two immediate questions are raised by the presence of a much thicker left-hand tail
of the firm size distribution. First, why do low productivity firms not improve their
productivity? Second, why do markets not force the less productive firms out of
business? Put simply, if management quality is important, why are some firms are
better managed than others? If access to export markets is critical for productivity,
why dont more firms export? If new high-yield seeds or mechanization dramatically
improve productivity in the agricultural sector, why dont more farmers adopt these
new seeds or technologies? We begin this discussion by presenting information on
the constraints perceived by the firms in the WBES. Specifically, in the WBES, firms
are asked to choose the largest obstacle they face from a list.
The exact wording of the question is as follows:
Figure 4: WBES Questionnaire
We use this to group constraints into the following categories:
Page 25 of 54
Institutions:
business licensing and permits; courts; customs and trade

regulations; labour regulations; tax administration; tax rates;
political instability; crime, theft and disorder; corruption
Infrastructure:
electricity, transportation
Finance:
access to finance
Land:
access to land
Other:
inadequately educated workforce, practices of the informal

sector
Figure 4A shows for each IGC country the main obstacle reported by small,
registered firms (firms with between 5 and 19 full-time employees.) To highlight
potential correlations with overall economic development, we order the countries by
2006 GDP per capita at PPP.
Page 26 of 54
Figure 4A: Main Obstacle for Small Firms
4.1 Credit Market Frictions
As can be seen, difficulty in accessing credit features prominently in all IGC

countries. For example, in Bangladesh close to 50% of small firms report finance as
being the key constraint on their expansion. Significant numbers of entrepreneurs
also report this constraint as being important in Ghana and Liberia. There is also
some evidence that suggests that firms in low income countries are constrained in
their access to finance9 and characterized by high average returns to capital10.
Broadly the evidence suggests that financial development is associated with higher
economic growth and lower poverty. The available evidence, however, has less to
say on what precise policy measures might relax credit constraints for firms. The
9
Burgess and Pande 2005, Banerjee and Duflo 2012, de Mel et al. 2008
10
de Mel et al. 2008, Fafchamps et al. 2011
Page 27 of 54
natural question is why such constraints would be binding in the long run if
entrepreneurs can build up their capital by saving.
There is also a growing body of work that suggests that small-scale interventions
designed solely to alleviate credit market frictions have little effect. For example,
many evaluations of microfinance suggest that access to credit alone does not help
create new businesses11. De Mel et al. [2010] show that subsidies to working capital
to allow firms to hire workers in Sri-Lanka had little effect. Specifically, the subsidy
was of a set amount that did not vary across firms and corresponded approximately
to 50% of the average low-skill wage in the region. Of the firms offered the subsidy,
only 22% employed a new worker.
A possible explanation is that credit market frictions are harmful in combination with
other constraints. For example, there is now a large body of evidence that standalone short-term training programmes are ineffective in increasing profits and
business growth12. Although there are promising exceptions13, it might be the case
that such training programmes are only effective if they are accompanied by
programmes to improve access to credit. For example, a recent study of business
training programmes for aspiring entrepreneurs with and without capital grants
provides evidence of such complementarity14. Bandiera et al (2013) also find that
when the poorest women in Bangladesh are offered both capital and training to
start up small rural enterprises they experience a 38 percent increase in earnings.
An explanation for the low take-up rate for the wage subsidy found in de Mel et al.
[2010] is that there are informational constraints in local labour markets. De Mel et
al. [2010] find that only 2% of the entrepreneurs in their sample placed any sort of
advertisement to find the new worker and that in as many as 85% of the cases,
entrepreneurs employed someone they already knew. But it might also be the case
that neither credit market failures nor business skills nor informational barriers are
the key constraints, and that other constraints are ultimately what matter.
11
Banerjee et al. 2010, Crepon et al. 2011, Karlan and Zinman 2011, Kaboski and Townsend 2011
12
Field et al. 2010, Karlan and Valdivia 2011, Fairlie et al. 2012, Bruhn et al. 2012, McKenzie and Woodruff 2012
13
Drexler et al. 2010
14
McKenzie and Woodruff 2012
Page 28 of 54
4.2 Institutional Constraints
An obvious candidate is constraints related to institutions. Indeed, in eight IGC

countries, more entrepreneurs report institutional constraints as being more
important than those relate to finance. In Zambia, Sierra Leone, Rwanda,
Mozambique and Liberia, it is the most frequently cited constraint from small firms.
Figure 4B shows the same information for large firms (those with at between 20 and
99 full time employees)
Figure 4B: Main Obstacle for Large Firms (size 20-99)
This highlights how the relative importance of constraints changes across the firm
size distribution. We see that financing constraints are generally less pressing for
these larger firms, while constraints related to institutions still feature prominently. In
Zambia, Sierra Leone and Rwanda such constraints remain the most frequent type
of constraints mentioned. For a number of IGC countries, smaller firms report such
Page 29 of 54
institutional constraints as significant obstacles to expansion and growth more

often than larger firms in the same country15
To get a more detailed picture of the underlying institutional constraints that
enterprises face, Figures 4C and 4D show the sub-components of institutional
constraints reported by small and large sized firms, across IGC countries. Again, we
again horizontally rank countries by GDP per capita in 2006 at PPP.
Figure 4C: Institutional Constraints for Small Firms (size 5-19)
15
Those countries in which institutional constraint are more frequently cited as an obstacle by small
firms than large include Liberia (small 41.9%, large 28.4%), Mozambique (small 29.4%, large 29.3%),
Rwanda (small 43%, large 39.4%), and Uganda (small 19.2%, large12.6%).
Page 30 of 54
Figure 4D: Institutional Constraints for Large Firms (size 20-99)
For small firms, taxes are seen as a major barrier to expansion. They dominate
reports in India, Zambia, Uganda, Sierra Leone, Rwanda and Ethiopia. They are also
seen as a prominent obstacle to expansion in large firms. This suggests that taxes
on firms, both small and large, may be preventing them from growing larger. This
may in part reflect the fact that taxes in developing countries have relatively narrow
bases, or are focused on tax bases such as the formal firm sector that are easy to
monitor and tax. Small informal firms in the service and manufacturing sector and
small farms may largely avoid taxation though, paradoxically, one of the key reasons
that these firms remain small may be that they want to avoid the punitive taxes that
would accompany firm growth.
For firms across the firm size distribution, corruption also features prominently as a
constraint on growth. There is a tendency for larger firms to report this more
frequently than smaller firms, although no clear pattern emerges between GDP per
capita and the percentage of firms reporting corruptions as a barrier to growth. This
Page 31 of 54
evidence is consistent with the body of evidence on such institutional constraints.

On corruption, Fisman and Svensson (2007) using self-reported bribery payments,
find that corruption has a strong negative effect on firm growth in Uganda. Cai et al.
(2005), using an alternative measure of bribery, also find that corruption had a
strong negative effect on firm performance in China, but this effect is much weaker
if firms are located in cities with low quality government services, if they are subject
to severe government expropriation and if they do not have a strong relationship
with clients and suppliers.
Labor regulation appears to constrain firm expansion in some countries with effects
being greater for larger firms. This is consistent with the available evidence. Besley
and Burgess (2004) exploit differences in labour regulation across Indian states and
over time to estimate their impact on firm performance. They find that more proworker regulations lower output, employment and investment in the registered
manufacturing sector. In contrast the find that pro-worker labor regulation increases
output in the informal sector as investors are more likely to remain small and
informal in states with more pro-worker regulations. Ahsan and Pages (2009) in the
same setting, distinguish between employment protection and disputes settlement
regulation. They find that the former has a negative impact on registered
manufacturing sector output and that this effect is larger in more labour intensive
industries.
Institutional constraints potentially have important effects on the internal
organisation of firms. For example, the difficulty in enforcing contracts and
punishing fraud may explain the lack of decentralisation documented by Bloom,
Sadun and Van Reenen (2009). The lack of decentralisation may also be related to
the high cost of information technology and to the absence of middle managers.
The link between trust and IT investments provides a useful link to the broader
literature on real versus formal authority16. Managers are more likely to delegate
down the hierarchy when the incentives of lower-level managers are aligned with
their own. The ability to provide direct incentives depends on information flows up
the organisation. For a variety of reasons some exogenous and some endogenous
to the institutional environment the cost of transmitting information across levels
of the hierarchy are likely to be higher in firms in LICs. The technology of information
collection is relatively expensive in LICs, and hence its adoption may be less
advanced. Further, thin markets for middle-level management limit the capacity to
16
Aghion and Tirole (1997)
Page 32 of 54
process the information which may be available.17 Decentralisation is also

associated with more intensely competitive markets18.
Yet, we note that in at least some low income countries, delegation does appear to
be possible. There is a large number of very large business groups involved in the
garments sector in Bangladesh and textiles in India, even though these are low trust
countries. To achieve this anecdotal evidence suggests firms do three things. First,
they try to adopt efficient monitoring systems so they can at least be aware of
malfeasance by their employees. If you are worried that employees may, for
example, be stealing raw materials from your firm then as least having extensive
monitoring limits the extent of this happening until you discover and reprimand the
employee. Second, they still largely operate within the confines of the family, so that
cousins, uncles, sons-in-law are often employed extensively, so try and exploit some
degree of family connection. Finally, to the extent that companies like Tata
decentralise decision making to outsiders they do this to long-serving and well paid
employees, who are effectively receiving a very high efficiency wage.
The institutional environment may also affect the nature of contracts between firms.
There is a greater reluctance in low-income countries to switch trading partners,
even for standardised products19. Information about reliability of firms and products
is poor and closely held. Formal legal remedies for contractual non-compliance are
slow, expensive and unreliable. As a result, buyer-seller relationships develop slowly
and buyers are reluctant to switch to new suppliers even when those suppliers offer
inputs at a lower price20. These contractual relationships are often managed by
market intermediaries. We have somewhat limited knowledge of the effect of these
intermediaries on producer firm dynamics. Much of the limited information we have
comes from descriptions of agricultural markets.
Such constraints may also affect the ability of a firm to invest in the products that
would allow them to grow. Diversification of exports, moving away from commodity
exports towards more sophisticated products, and upgrading the quality of existing
exports feature prominently on the policy agenda of many developing countries.
Hausman, Rodrik and Wang (2005) find evidence that exports of more-sophisticated
products are positively correlated with subsequent growth. But this is far from a
17
Here we rely on ongoing IGC-sponsored work by Woodruff and MacChiavello with large garment manufacturers in
Bangladesh.
18
McAfee and McMillan 1995; Guadalupe and Wulf 2008
19
See, for example, Johnson, McMillan and Woodruff 2002
20
Banerjee and Duflo 1999; McMillan and Woodruff 1999; MacChiavello 2013
Page 33 of 54
settled matter as some argue that there is no clear case in favour of this view. We
believe this issue to be of enormous importance. We will support research that
takes advantage of the newly available firm-product-level information to investigate
further how product composition evolves as firms learn and grow, and how this
process relates to trade patterns and growth at the country level.
This focus includes agriculture as well as manufacturing: agricultural producers
seeking to expand into non-traditional products or upgrade existing processes and
products face many of the same challenges as manufacturing firms. For an example
of this kind of investigation, see Lederman, Rodriguez-Clare and Xu (2011), who
investigate how export growth in Costa Rica over the period 1997-2007 is explained
by expanding exports of the same products by the same firms to the same
destinations versus the reshuffling and reallocations among firms, products and
destinations (including entry and exit) over time. Of interest also is a recent
compendium of case studies in Latin America of the birth of new export sectors21.
Figure 4E brings to the fore how the ranking of constraints differs by firm size in the
same country (a rank of one implies it is the most frequently named type of
institutional constraint). For most constraints, we see a high correlation between the
importance of that constraint as reported by small and large firms. The one notable
exception is taxes: this is a far higher ranked constraint for large firms rather than
small firms in the same economy.
21
Fernndez-Arias et al (2012)
Page 34 of 54
Figure 4E: Ranking of Institutional Constraints by Firm Size
The differing constraints for big versus small firms is important in thinking about the
underlying sources of the skewed size distribution and for the large degree of
informality. For example, De Sotos (1989) work has persuaded many policymakers
that many firms want to become formal firms but burdensome regulation prevents
small firms from formalising. Partly in response to these arguments, nearly threequarters of the countries included in the World Bank and IFCs Doing Business
surveys have adopted reforms designed to facilitate business registration22.
However, the evidence from such reforms suggests that reducing the costs of
formalising has little effect. Much of the work in this area has been done in Mexico,
which has both high-quality data on informal firms via its Economics Census and
has enacted significant reforms to the formalisation process, including reducing the
time to register a firm in some sectors from 30 to 2 days at the municipal level.
22
IFC, 2009
Page 35 of 54
These reforms induced modest increases in formalisation (the fraction of registered

businesses increased by 5%) but the mechanism through which even these small
changes were affected remains uncertain. Bruhn (2011) finds that any increases
appear to be due to new entry rather than the formalisation of existing firms, while
Kaplan et al. (2011) use different data and find the opposite result. Furthermore,
Busso, Fazio and Levy (2012) provide clear evidence that the reality is much more
complex than simply whether a firm is formal or not. They show that in Mexico, most
firms are in fact formally registered but still remain small because they can evade
taxes by remaining small.
The evidence from the WBES presented above is consistent with the evidence in
Busso, Fazio, and Levy (2012). Specifically, the fact that the constraints differ
between small versus large firms suggests that informality and the skewed size
distribution may be driven by the perceived benefits of informality (and remaining
small) rather than the high upfront cost of formalisation. Put simply, the tradeoff is
between the importance of the constraints that big firms face versus the constraints
that small firms face, which ultimately is an empirical question.
There is a growing body of evidence that the skewed size distribution and
informality stems from benefits of informality relative to the benefits of formality.
Maloney (2004) suggests that smaller and less productive firms may get little benefit
from formalization, and argues that many firms may rationally choose to remain
small and informal. McKenzie and Sakho (2010) demonstrate significant
heterogeneity in response to firm registration in Bolivia. Even if formalisation may
bring benefits to some small firms, perhaps these benefits are not universal.
Jaramillo [2013] studies a registration subsidy for microenterprises in Lima, Peru. He
finds that the limited growth aspirations of these firms combined with the recurring
costs and low perceived benefits of formalisation yield very low demand for
formalisation.
De Mel et al. (2013) provide clear evidence that reducing formalisation costs alone
may have a limited impact on registrations and provide further support for the
hypothesis that a lack of meaningful ongoing benefits may be the key obstacle to
increasing formalisation. This study builds on a randomised experiment testing
different incentives for formality in Sri Lanka. Providing information about the
registration process and reimbursing all direct costs was not enough to get firms to
register. Even with powerful monetary incentives, up to an additional two months of
the median profits for firms in their sample, only half of firms choose to register.
Page 36 of 54
4.3 Infrastructure
Infrastructure is seen as a major impediment to expansion and growth in most

developing countries and this is true for both small and large firms. In Figures 4A
and 4B we see that infrastructure is seen as more important than finance and
institutions for a number of countries. For some IGC countries such as Tanzania and
Uganda, infrastructure is the dominant constraint. For Liberia, land is the most
binding constraint for larger firms. There is of course a large body of evidence on
the importance of infrastructure. Aschauer (1989) found a positive relation between
infrastructure capital and TFP in the United States. Mitra et al. (2002) estimate this
effect for Indian manufacturing sector. They find that infrastructure endowments
explain a large part of TFP differences across Indian states. Since the effect they
estimate is larger than the one found for the United States, they argue that majority
of states in India face a strong infrastructure bottleneck. Donaldson (forthcoming)
provides convincing evidence that the expansion of railways in India during the
colonial period had a large effect in terms of raising per capita agricultural income.
Within this literature, another strand has focused on specific components of
infrastructure and their impact on economic outcomes. For example, in the area of
energy, Reinikka and Svensson (1999) use firm-level data to show that the lack of
reliable power supply in Uganda reduces private investment productivity by forcing
firms to invest in generators and other low-productivity substitutes for reliable public
provision of power. The magnitude of these effects remains in question. Recent
work by Alcott et al (2013) studies the impact of power shortages on firms
productivity in India. They find that although power cuts are perceived as very
damaging by entrepreneurs, the estimated effect is relatively small.
The scope of the market faced by firms may also be limited by the high costs of
transportation arising from under-developed transportation networks. There is a
recent burst of work related to the effects of transportation infrastructure on a
variety of outcomes, including market integration and firm dynamics23. In IGG
sponsored work, for example, Casaburi, Glennerster and Suri (2013) use a
regression discontinuity effect to show that road and bridge expansion following
civil conflict in Sierra Leone improved market integration and lowered crop prices by
improving the match between buyers and sellers. Given the scale of investments in
infrastructure, a more careful mapping of the effects of transportation costs on
23
See, for example, Donaldson 2012 for evidence of transportation cost on the integration of
markets and Ghani et al 2012 for more direct evidence on transportation networks and firm dynamics
Page 37 of 54
competition is needed. In particular, the fragmentation of markets caused by the

lack of transportation means that highly productive firms in one market struggle
to take market share from less productive firms in other markets.
Infrastructure bottlenecks clearly potentially limit the amount of domestic and
international trade. While there is a long body of research of barriers to international
trade, there is increasing evidence that trade costs internal to countries are
significant and contribute to marked differences in prices, incomes and growth
across regions. But we know little about the nature of these internal trade costs and
how reducing them would affect economic outcomes. How large are such costs?
Are they systematically higher in poor countries? What are their key determinants?
To what extent would poor regions benefit from increased investment in roads and
other transport infrastructure? Answers to this latter question will shed new light not
only on how infrastructure investments should be prioritised domestically, but also
potentially on the consequences of integration between rich and poor countries.
4.4 Skills and Training
Another important constraint facing firms in low income countries is the absence of
adequately skilled workers and managers. Bloom et al (2013), for example, provide
anecdotal evidence of the difficulty of finding skilled managers in India. One
response of the policy community to this perceived failure are training programmes.
Attanasio et al. (2011) use a randomised design to study Jovenes en Accion, a
training programme targeting disadvantaged youth in Colombia. Card et al. (2011)
evaluate a very similar training programme for disadvantaged youth in the
Dominican Republic. These is some evidence that these programmes had small
effects on labour force participation rates and wages, but there is no evidence that
such programmes have a long run effect on occupational structure and questions
remain as to whether these programmes crowd out other mechanisms through
which people acquire skills. In addition, the goal of all of these programmes is to
increase the supply of workers with basic skills, and it is not clear whether this is the
binding constraint in terms of labour supply or whether the constraint relates to the
scarcity of skilled managers24.
It could also be the case that the most important vehicle through which workers
obtain skills is via on-the-job training. The World Bank Enterprise Survey data allows
us to explore related issues as measured in the surveyed firms (that covers only
24
As suggested by Bloom et al
Page 38 of 54
formal firms) and among currently employed workers. Figure 5A shows the
percentage of firms reporting to provide training to their workers, by IGC country25:
Figure 5A: Provision of Formal Training
As can be seen, many firms provide training and this tendency is more pronounced
among larger firms. The WBES data refers to formal training but on the job
training may also be important, and there are reasons to believe that the workers in
large formal firms may also get more on the job training. So here, the dominance
of small firms in aggregate employment may also have the effect that many workers
may not learn new skills that would allow them to increase their wages or to start a
new business. We do not know the importance of this, and this is work we want to
encourage in the future.
25
Ranked by GDP per capita in 2006
Page 39 of 54
4.5 Firm Clusters
Moving beyond the material captured in available firm level datasets, we have
abundant evidence that the clustering of firms, particularly in industrial sectors,
produces positive externalities that are important for capturing local spillovers and
fostering industrial development. One of the important spillover effects of this is in
knowledge: firms may learn from their neighbours, either directly or by hiring
workers from nearby firms. Careful research has been done to advance our
understanding of such externalities, but the theoretical and econometric challenges
remain formidable, especially within developing countries. One key issue is the
extent to which knowledge spillovers occur from investment by multinational
corporations. There is considerable scope for doing research, for example, on how
foreign direct investment and trade liberalization promote firm clustering and
expand the span of products that firms can produce (see Goldberg et al, 2010).
Much of the evidence we have on where and why firm clusters form in particular
parts of developing countries is descriptive and anecdotal. There is much more
research that could be done in this important area.
The existence of externalities opens up the possibility of welfare-increasing
government interventions. There is however little in the way of rigorous empirical
evaluation of such policies. Evaluating industrial policies is challenging since they
differ widely across sectors and countries and there is typically no clear
counterfactual for comparison. This is especially true in the case of what is often
called new industrial policy, which advocates a minimalist approach in which
governments avoid price distortions and instead establish a process to discover
bottlenecks and coordination failures and act directly to resolve them.
4.6 Markets for Inputs and Outputs
The structure of the market for inputs and outputs may also play an important role in
addition to its incentive effects on productivity. In the IGCs previous work in Sierra
Leone,
Rwanda, and Pakistan, our research has shown that the market structure of the
agricultural supply chain has important effects on prices and thus farmer welfare
(see e.g. Casaburi, Glennerster and Suri 2013). IGC research in Ghana and India
also shows that the availability of insurance led farmers to invest more and also
served to establish the existence of a demand for agricultural insurance in some
regions. We suspect that other critical inputs are also missing and that similar forces
are at play in the industrial sector.
Page 40 of 54
Specifically, uncompetitive input markets may increase prices and make critical
inputs scarce, which have large potentially adverse effects on firm performance. In
the agricultural sector, inefficient firms in sectors that sell seeds and mechanical
farming equipment could explain why so many farmers still use primitive
technologies. One of the key reasons why moving agriculture closer to the
production frontier can be beneficial is that the surpluses so generated are often
invested in human capital acquisition, migration, and small business creation, which
themselves promote economic diversification and economic growth. In the industrial
sector, the inefficient electricity sectors in India and Pakistan (and many other
countries), the absence of functioning markets for land and office space in urban
areas, and the inefficient transportation sector might be behind why many firms
remain small and unproductive. Again, more research is required to accumulate the
necessary systematic evidence to rigorously approach these questions.
Page 41 of 54
5. Resource Allocation
The last pillar of the research on firm capabilities encompasses understanding the
importance of and barriers to effective resource allocation. There exists suggestive
evidence that growth is driven by the entry and emergence of new firms and sectors
as well as the reallocation of resources from less productive to more productive
firms. If resources do not flow into more productive new firms and sectors and if
there are impediments to resources being reallocated from unproductive firms and
sectors to more productive areas the economic growth is likely to be low for a
sustained period of time. The study of firm capabilities therefore has to encompass
the study of firm dynamics both in terms of entry and growth of new firms but also
in terms of whether more productive firms are growing and less productive ones
dying out.
The research by John Sutton funded by the IGC shows has been important in terms
of identifying where large firms which produce and export manufactured goods
came from in Africa. His finding they originated as trading firms rather than as small
producers of manufactures suggests that there may be significant difficulties for
small manufacturing firms to grow to scale in these countries. New industrial
sectors, such as the flower sectors in Kenya and Ethiopia, are particularly interesting
to study as they point to how impediments to resource reallocation can be
overcome. IGC has done some path breaking working in this area focussed on how
flower producers in Kenya overcome contracting problems with purchaser of
flowers in international markets (see MacChiavello and Morjaria 2013).
We would like to see much more of type of work where the focus is on identifying
constraints on resource reallocation and trying to identify how these constraints can
be overcome. More broadly the IGC Firm Capabilities Programme would like to
understand the importance of the reallocation of resources from informal to formal
firms in manufacturing, from non-exporters to exporters, from small family owned
farms to large corporate farms, and from agriculture to industry. Studying these
processes is challenging as it involves looking at large populations of firms across
time but the returns, in terms of understanding where growth comes from in
developing countries and how it can be encouraged by institutional and policy
interventions, is also large.
The available evidence we have suggests that the necessary resource reallocation
does not take place in many developing countries. This results in industrial sectors
Page 42 of 54
being both small and dominated by a large number of small, unproductive, informal
firms. In the agricultural sector, the vast majority of land and people are employed in
small subsistence farms, despite the presence of potentially more productive farms
producing cash crops with modern technologies. Workers and capital remain locked
in less productive agricultural sectors, while it may be the case that these resources
could be more profitably employed in other sectors either in industry or in
commercial agriculture.
These observations throw up a number of questions which are fundamental to
understanding economic growth in developing countries. What are the barriers that
prevent the efficient reallocation of resources? What prevents subsistence farmers
from moving into cash crops or into urban industrial sectors? We have a long list of
candidate explanations, but there is currently little evidence of which explanations
hold the most empirical water. The research undertaken in the coming years will
serve to shed light on these questions.
Perhaps nowhere is the issue of resource allocation more important than in trade.
There is a large consensus that trade is critical in the development of firm
capabilities, and more broadly in allowing poor countries to catch up to rich
countries. There are many well-established theories that outline the mechanisms
through which this can happen. Traditional theories emphasize how trade allows
countries to specialise in sectors where they have comparative advantage26, while
new theories emphasise reallocations of resources across firms within industries27,
across products within firms and across destination markets for sales of particular
products within firms28. Trade allows developing countries to import sophisticated
new technology embodied in capital and intermediate goods29. Trade may also lead
to gains from increased competition, which may increase allocative efficiency30.
There is a widespread perception that the effects of trade on the development
process goes well beyond static reallocations. In particular, trade is thought to
affect the growth process through its impact on the accumulation of knowledge
knowledge about how to produce existing products at lower cost or higher quality,
about how to produce new products, about what a country is good at producing,
about where to find suppliers or buyers, or about what customers want.
26
See, for example, Eaton and Kortum, 2002
27
Melitz, 2003
28
Bernard, Redding and Schott, 2010
29
Eaton and Kortum, 2001
30
Midrigan and Xu, 2012, and Holmes, Hsu and Lee, 2013
Page 43 of 54
Yet the question is why many countries remain poor despite the apparent potential
to grow via trade. One hypothesis is that resources need to be reallocated towards
sectors in which they have a comparative advantage for this to take place. A related
hypothesis is that poor countries may not always reap the benefits of trade
liberalisation because of the absence of complementary policies and institutions. As
support for these hypotheses, Menezes-Filho and Muendler (2011) track individual
workers across jobs after Brazil's trade liberalisation in the 1990s and find that "tariff
cuts trigger worker displacements, but neither exporters nor comparative advantage
sectors absorb trade-displaced labour." They conclude that, contrary to standard
theory, trade liberalisation did not lead to a reallocation of workers out of importcompeting sectors and into exporting sectors, but rather into services or
unemployment or simply out of the labour force.
One key impediment to labour moving into more productive sectors may be labour
regulations. Labour regulations make it unattractive for high productivity firms to
expand their employment levels, and this may lead these firms instead opt for
employing more capital. Some evidence of this is shown in Aghion, Burgess,
Redding and Zilibotti (2008), who show how the dismantling of the License Raj in
India (a system of controls affecting entry and production) had more positive
effects on manufacturing output in states with more pro-employer labour market
institutions. In states where hiring and firing workers was easier labour was more
able to flow towards more productive firms and this, in turn, raised labour
productivity more than in states where getting rid of unwanted workers was more
difficult. Firms in states with pro-employer regulations thus benefited more from the
delicensing liberalization than firms in states with pro-worker regulations. This
suggests that the growth impact of trade liberalization will be lower in countries
where labour market regulations impede the movement of workers across firms.
Barriers to resource allocation are also important in thinking about the sectoral
composition of the economy. Recent research shows that poor countries are
particularly unproductive in agriculture. For example, Tombe (2012) finds that
average labour productivity in agriculture differs by a factor of 70 between the
poorest and richest countries, but by only a factor of 6 in non-agriculture. This
implies that poor countries have a comparative advantage in non-agriculture
tradable sectors, i.e., manufacturing, and yet they devote a large fraction of their
resources to agriculture. In a closed economy, this can be explained by nonhomothetic preferences: being poor, households in poor countries must spend a
large fraction of their income on food, hence leading to a large fraction of labour to
Page 44 of 54
be employed in agriculture. But countries can trade, and so the relevant question is
why do they not import food and specialise in manufacturing? Presumably trade
costs are large, both to trade with the outside world because of bad port
infrastructure and shipping connections (or landlocked countries), and because of
poor infrastructure connecting ports to the inland markets relevant for households31.
31
Tombe (2012)
Page 45 of 54
6. Conclusion
Relative to the study of consumption and households, the study of production and
firms has received little attention within development economics. This is changing
quickly as macroeconomists and economists interested in industrial development,
trade, agricultural development and entrepreneurship move into the study of firm
capabilities in developing countries. The IGC Firm Capabilities Research
Programme pulls economists with this common interest in firm capabilities together
to focus on three core questions. (i) What are the key proximate determinants of
firm productivity? (ii) Where does the productive capacity of firms comes from? (iii)
What are the barriers that prevent resources from moving from unproductive firms
and sectors to areas of higher productivity?
One key finding from this paper is that we know staggeringly little about the
production sector in developing countries. Part of the challenge of the IGC Firm
Capabilities Research Programme will therefore be to gather better data on firm
productivity across the full firm size distribution in developing countries. By
gathering this information over time and in a larger number of countries we will be
better able to study both the drivers of productivity as well as the resource
allocation and structural change processes that underpin economic growth.
Building up panel data sets where large numbers of firms can be followed over time
is a particular priority as this would allow us to look at firm dynamics and to examine
both how the entry of products and firms and the reallocation of resources from
unproductive to productive firms drive economic growth. Given the structure of
undeveloped economies the study of resource reallocation must encompass the
movement of workers from subsistence agriculture into industry and of workers in
informal industrial sectors into formal industrial sectors. Linking household data sets
with firm data sets over time will be useful in this respect as individuals often divide
their time between household based production and outside employment. These
processes of occupational and structural change have been understudied in recent
years but are fundamental to understanding where economic growth comes from in
developing countries.
Our review of the available data on firms in developing countries throws up three
major puzzles. The first is why the size of the industrial sector remains so small
within the overall economy? This leads us to focus on industrial development as key
policy objective in developing countries. The evidence base for discussing industrial
Page 46 of 54
policy has been strikingly poor in developing countries so there a real opportunity to
bring new data and theory to bear on this issue. Understanding the role of trade and
competition in driving the growth of industrial sectors is likely to be particularly
important here. Whether or not firms can effectively respond to greater competitive
pressures which come from opening up parts of the economy and whether or not
foreign direct investment and international trade will bring in new knowledge and
allow new products to be produced are both major questions in this area.
The second puzzle concerns why production is concentrated within firms in the left
hand tail of the productivity distribution? This leads us to a focus on understanding
what factors are behind differences in firm productivity that we observe both within
and across countries. We have reviewed information related to credit market
frictions, institutional constraints, infrastructure, skills and training, firm clusters and
markets for inputs and outputs. As our review makes plain most of the evidence we
have to hand is not rich enough to base economic policy on. We can discern
relationships between labour productivity and some policy variables in the data but
we are far from understanding what types of interventions will unlock the productive
potential of firms. This is, in part, because the set of factors that keep firms in a low
labour productivity trap are complex and likely to interact with one another.
Providing capital to a small manufacturing firm in Liberia, for example, may have not
any discernible impact on labour productivity if the firm cannot find skilled workers.
And the need to rely on diesel power generation because of frequent power
outages may mean the products the firm produces cannot be competitive to world
markets. If one adds the need to pay off industrial inspectors and customs officials
then the difficulty of increasing productivity becomes apparent. Attacking single
constraints in a disjointed way is unlikely to work.
The third puzzle is why more resources not reallocated from unproductive firms and
sectors to productive firms and sectors in developing economies? This leads us to
focus on structural transformation and structural change within an economy and to
better understand why these change processes are so slow in many developing
countries. We have discussed some factors such as regulations and trade barriers
which may constrain the ability of factors of production to move across firms and
sectors within an economy. But it is clear that we are only scratching the surface.
We are at a point where we can describe and quantify the extent to which an
improved allocation of resources would contribute to productivity growth but are
unable to say how we would go about implementing such a reallocation. Getting
into the detail of this will require tracking firms over time and getting into the detail
Page 47 of 54
of how different policy interventions affect resource reallocation. This will be a major
challenge going forward.
Production in developing countries is concentrated in firms that are both small and
unproductive. This is true whether we look at agriculture, manufacturing or services.
The small size of the industrial sector and the focus of production on unproductive
firms helps to explain why aggregate productivity is so low in developing countries.
And low productivity ultimately explains why living standards are so much lower in
these countries. Understanding why firm capabilities are so low in developing and
uncovering ways both to increase labour productivity and to move resources from
unproductive firms and sector to productive ones are therefore of fundamental
importance not just for understanding what drives economic growth in developing
countries but also for finding the means to close the gap in living standards
between developing and developed nations.
Page 48 of 54
References
Aghion, Philippe and Jean Tirole (1997), Formal and Real Authority in
Organizations, Journal of Political Economy, 105, pp. 1-29
Aghion Philippe, Robin Burgess, Stephen J. Redding and Fabrizio Zilibotti (2008),
The Unequal Effects of Liberalization: Evidence from Dismantling the License Raj in
India, American Economic Review, 98(4), pages 1397-1412
Ahsan Ahmad and Carmen Pages (2009), Are All Labour Regulations Equal?
Evidence from Indian Manufacturing, Journal of Comparative Economics 37, no. 1:
62-75.
Aschauer David (1989), Is Public Expenditure Productive? Journal of Monetary
Economics 23: 177200.
Attanasio Orazio, A. Kugler and Costas Meghir (2011), Subsidized Vocational
Training for Disadvantaged Youth in Developing Countries: Evidence from a
Randomized Trial, American Economic Journal: Applied Economics 3: 188-220.
Bandiera, Oriana, Luigi Guiso, Andrea Prat and Raffaella Sadun (2011), What Do
CEOs Do? Harvard Business School Working Paper 11-081
Bandiera Oriana, Robin Burgess, Narayan Das, Selim Gulesci, Imran Rasul and
Munshi Sulaiman (2013), Can Basic Entrepreneurship Transform the Economic
Lives of the Poor? IZA Discussion Papers 7386, Institute for the Study of Labour
(IZA).
Banerjee Abhijit, Esther Duflo, Rachel Glennerster and Cynthia Kinnan (2010), The
Miracle of Microfinance? Evidence from a Randomized Evaluation, mimeo MIT.
Banerjee Abhijit, and Esther Duflo (2012), Do Firms Want to Borrow More? Testing
Credit Constraints Using a Directed Lending Program, mimeo MIT.
Banerjee Abhijit, Esther Duflo, R. Chattopadhyay and Jeremy Shapiro (2012),
Targeting the Hardcore Poor: An Impact Assessment, mimeo MIT.
Bernard, Andrew B., Stephen Redding and Peter Schott (2010), Multiple Product
Firms and Product Switching, American Economic Review, 100(1):70-97
Page 49 of 54
Besley Timothy and Robin Burgess (2004), Can Labour Regulation Hinder
Economic Performance? Evidence from India, Quarterly Journal of Economics 119,
no. 1: 91-134.
Bloom, Nicholas, Benn Eifert, Aprajit Mahajan, David McKenzie and John Roberts
(2013), Does management matter: evidence from India, Quarterly Journal of
Economics 128(1)
Bloom, Nicholas, Raffaella Sadun and John Van Reenen (2012), The organization of
firms across countries, Quarterly Journal of Economics, 127(4):1663-1704
Bloom, Nicholas and John Van Reenen (2010) Why Do Management Practices
Differ Across Firms and Countries?, Journal of Economic Perspectives, 203224
Bruhn Miriam (2011), License to Sell: The Effect of Business Registration Reform on
Entrepreneurial Activity in Mexico, Review of Economics and Statistics 93 (1): 382
86.
Bruhn Miriam, Dean Karlan and Antoinette Schoar (2012), The Impact of
Consulting Services on Small and Medium Enterprises: Evidence from a
Randomized Trial in Mexico, mimeo Yale University.
Burgess, Robin and Rohini Pande (2005), Do Rural Banks Matter? Evidence for the
Indian Social Banking Experiment? American Economic Review 95(3)
Busso Matias, Maria Victoria Fazio and Santiago Levy (2012), (In)Formal and
(Un)Productive: The Productivity Costs of Excessive Informality in Mexico,
Research Department Publications 4789, Inter-American Development Bank,
Research Department.
Cai Hongbin, Hanming Fang, Lixin Colin Xu (2005), Eat, Drink, Firms and
Government: An Investigation of Corruption from Entertainment and Travel Costs of
Chinese Firms, NBER Working Papers 11592, National Bureau of Economic
Research, Inc
Card David, Pablo Ibarran, Ferdinando Regalia, David Rosas-Shady and Yuri Soares
(2011), The Labour Market Impacts of Youth Training in the Dominican Republic,
Journal of Labour Economics 29: 267-300.
Page 50 of 54
Casaburi Lorenzo, Rachel Glennerster and Tavneet Suri (2013), Rural Roads and
Intermediated Trade: Regression Discontinuity Evidence from Sierra Leone, mimeo
MIT.
Crpon Bruno, Florencia Devoto, Esther Duflo and W. Parient (2011), Impact of
Microcredit in Rural Areas of Morocco: Evidence from a Randomized Evaluation,
mimeo MIT.
de Mel Suresh, David McKenzie, and Chris Woodruff ( 2013), The Demand for, and
Consequences of, Formalization among Informal Firms in Sri Lanka, American
Economic Journal: Applied Economics, American Economic Association, vol. 5(2),
pages 122-50, April.de Mel Suresh, David McKenzie, and Chris Woodruff (2010),
Wage Subsidies for Microenterprises, American Economic Review, 100(2): 61418. de Mel Suresh, David
McKenzie, and Chris Woodruff (2008), Returns to Capital in Microenterprises:
Evidence from a Field Experiment, Quarterly Journal Of Economics 123, no. 4:
1329-1372.
De Soto Hernando (1989), The Other Path: The Economic Answer to Terrorism. New
York: Harper and Row Publishers.
Donaldson Dave (forthcoming), "Railroads of the Raj: Estimating the Impact of
Transportation Infrastructure," American Economic Review.
Donaldson Dave, Arnaud Costinot and Ivana Komunjer (2012), What Goods Do
Countries Trade? A Quantitative Explanation of Ricardos Ideas, Review of
Economic Studies, 79(2): 581-608
Drexler Alejandro, Greg Fischer and Antoinette Schoar (2010), Keeping It Simple:
Financial Literacy and Rules of Thumb, CEPR Discussion Paper 7994Eaton,
Jonathan and Samuel Kortum (2001), Trade in capital goods, European Economic
Review 45 (7), 11951235
Eaton, Jonathan and Samuel Kortum (2002), Technology, Geography and Trade,
Econometrica 70(5), 17411779
Page 51 of 54
Fairlie, Robert, Dean Karlan and Jonathan Zinman (2012), Behind the gate
experiment: Evidence on effects of and rationales for subsidized entrepreneurship
training, Working Paper.
Fafchamps Marcel, David McKenzie, Simon R. Quinn and Chris Woodruff (2011),
When is capital enough to get female microenterprises growing? Evidence from a
randomized experiment in Ghana, NBER Working Papers 17207, National Bureau of
Economic Research.
Field Erica, Seema Jayachandran and Rohini Pande (2010), Do Traditional
Institutions Constrain Female Entrepreneurial Investment? A Field Experiment on
Business Training in India, American Economic Review Papers and Proceedings:
100:125-29.
Foster, Lucia, John Haltiwanger and Chad Syverson, (2008) Reallocation Firm
Turnover, and Efficiency: Selection on Productivity or Profitability? American
Economic Review, 98, 1:394 425.
Foster, Lucia, John Haltiwanger and C. J. Krizan (2006) Market Selection,
Reallocation, and Restructuring in the U.S. Retail Trade Sector in the 1990s, Review
of Economics and Statistics 88(4):748-758
Goldberg, Penelope, Amit Khandelwal, Nina Pavcnik, and Petia Topolova (2010),
"Imported Intermediate Inputs and Domestic Product Growth: Evidence from India,"
Quarterly Journal of Economics 125(4): 1727-1767.
Hausmann Ricardo, Jason Hwang and Dani Rodrik (2005), What You Export
Matters, CID Working Paper 123, Center for International Development, Harvard
Holmes, Thomas, Wen-Tai Hsu and Sanghoon Lee (2013), Allocative Efficiency,
Mark-ups, and the Welfare Gains from Trade, Working Paper
Hsieh, Chang-Tai and Peter Klenow (2009), Misallocation and Manufacturing TFP
in China and India, Quarterly Journal of Economics
Hsieh, Chang-Tai and Peter Klenow (2012), "The Life Cycle of Plants in India and
Mexico," mimeo University of Chicago
International Finance Corporation (IFC) (2009), Doing Business 2010: Reforming
through difficult times. IFC: Washington, D.C.
Page 52 of 54
Jensen Robert (2012), Do Labour Market Opportunities Affect Young Women's

Work and Family Decisions? Experimental Evidence from India, Quarterly Journal of
Economics 127: 753-92.
Kaboski Joseph and Robert M. Townsend (2011), A Structural Evaluation of a
Large-Scale Quasi-Experimental Microfinance Initiative, Econometrica 79: 1357406.
Kaplan David, Eduardo Piedra, and Enrique Seira (2011), Entry Regulation and
Business Start-Ups: Evidence from Mexico, Journal of Public Economics 95 (11
12): 150115.
Karlan, Dean and Martin Valdivia (2011), Teaching Entrepreneurship: Impact of
Business Training on Microfinance Clients and Institutions, Review of Economics
and Statistics 93: 510-27.
Karlan, Dean and Jonathan Zinman (2011), Microcredit in Theory and Practice:
Using Randomized Credit Scoring for Impact Evaluation, Science 332: 1278-84.
Lederman, Daniel, Andres Rodriguez-Clare and Daniel Yi Xu (2011),
Entrepreneurship and the Extensive Margin in Export Growth: A Microeconomic
Accounting of Costa Ricas Export Growth during 1997-2007, World Bank
Economic Review, 25:3
MacChiavello, Rocco and Ameet Morjaria (2013), The Value of Relationships,
American Economic Review (R&R)

Maloney William F. (2004), Informality Revisited. World Development 32 (7): 1159
78.
McKenzie, David and Chris Woodruff (2012), What Are We Learning from Business
Training and Entrepreneurship Evaluations Around the Developing World? mimeo,
Warwick University.
McKenzie David and Yaye Seynabou Sakho (2010), Does It Pay Firms to Register
for Taxes? The Impact of Formality on Firm Profitability, Journal of Development
Economics 91 (1): 15 24.
Melitz Marc (2003), The Impact of Trade on Intra-Industry Reallocations and
Aggregate Industry Productivity, Econometrica, 71(6):1695-1725
Page 53 of 54
Menezes-Filho, Narcio Aquino Menezes Filho, Marc-Andreas Muendler (2011),

Labor Reallocation in Response to Trade Reform, NBER Working Paper 17372,
National Bureau of Economic Research
Midrigan, Virgiliu, Chris Edmond and Daniel Yi Xu (2012), Competition, Markups
and the Gains from International Trade, Working Paper
Mitra Arup, Aristomene Varoudakis and Marie-Ange Vganzons-Varoudkis
(2002), Productivity and Technical Efficiency in Indian States Manufacturing: The
Role of Infrastructure, Economic Development and Cultural Change 2002: 395
426.
Reinikka Ritva and Jakob Svensson (1999), How Inadequate Provision of Public
Infrastructure and Services Affect Private Investment? Policy Research Working
Paper 2262, World Bank, Washington, DC.
Tombe Trevor (2012), The Missing Food Problem, Working Paper
Page 54 of 54

I GC Evidence Paper Firms

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

I GC Evidence Paper Firms

Uploaded by

Copyright:

Available Formats

EVIDENCE PAPER

FIRM CAPABILITIES AND

The International Growth Centre (IGC) aims to promote sustainable

Evidence Paper: Firm Capabilities and Economic Growth

Stylised Facts ................................................................................................................. 12

Determinants of Low Labour Productivity ..................................................................... 25

Credit Market Frictions ........................................................................................... 27

Institutional Constraints .......................................................................................... 29

Skills and Training ................................................................................................... 38

Firm Clusters ........................................................................................................... 40

Markets for Inputs and Outputs ............................................................................. 40

Resource Allocation ....................................................................................................... 42

Evidence Paper: Firm Capabilities and Economic Growth

Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

employed in small subsistence farms, despite the presence of potentially more

Evidence Paper: Firm Capabilities and Economic Growth

unproductive firms and sector to productive ones are therefore of fundamental

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

All but Myanmar and South Sudan

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

breakdown in sectoral composition focuses on firms with 0-49 full-time employees,

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Evidence Paper: Firm Capabilities and Economic Growth

Figure 3A: Firms Formally Registered When Started

Evidence Paper: Firm Capabilities and Economic Growth

Figure 3B: Number of Years Firms Operated Without Registration

Evidence Paper: Firm Capabilities and Economic Growth

Figure 3C: Sectoral Composition in Mexico, by Establishment Size

Yet as shown in Figure 3B these differences persist, albeit in smaller magnitude,

Evidence Paper: Firm Capabilities and Economic Growth

Figure 3D: Firm Size Distribution by Sector in Mexico, Excluding Establishments of

Evidence Paper: Firm Capabilities and Economic Growth

entrepreneurs engaged as brokers in foreign trade. We want to know whether this

Evidence Paper: Firm Capabilities and Economic Growth

4. Determinants of Low Labour Productivity

We use this to group constraints into the following categories:

Evidence Paper: Firm Capabilities and Economic Growth

business licensing and permits; courts; customs and trade

inadequately educated workforce, practices of the informal

Evidence Paper: Firm Capabilities and Economic Growth

Figure 4A: Main Obstacle for Small Firms

4.1 Credit Market Frictions

As can be seen, difficulty in accessing credit features prominently in all IGC

de Mel et al. 2008, Fafchamps et al. 2011

Evidence Paper: Firm Capabilities and Economic Growth

Drexler et al. 2010

McKenzie and Woodruff 2012

Evidence Paper: Firm Capabilities and Economic Growth

4.2 Institutional Constraints

An obvious candidate is constraints related to institutions. Indeed, in eight IGC

Evidence Paper: Firm Capabilities and Economic Growth